He won 67% of the time. His average loss was bigger than his average win. The same ledger says both.

We measured this from the wallet file he published: 2,172 realised-PnL events recorded by the exchange between March 2018 and December 2021.

1,455 wins, 717 losses. Win rate 67.0%.

Average win: 5.90 BTC. Average loss: 7.04 BTC.

That is a payoff ratio of 0.84. Below one. The opposite of the rule every course sells.

Look at the medians and it flips: median win 0.64 BTC, median loss 0.54 BTC, ratio 1.17.

So the typical trade was fine. A small number of very large losses dragged the average down.

The edge is in what survives that: profit factor 1.70, and +1.63 BTC expected per event.

By day: 1,379 days with realised PnL, 66.1% of them green. Average green day +8.95 BTC, average red day -9.90 BTC.

His best and worst days are both crashes. Best: 13 March 2020, +275.5 BTC. Worst: 20 May 2021, -281.8 BTC.

2021 is the strangest year in the file. Win rate 71.4%, the highest. Payoff ratio 0.70, the worst. Net result +2,069 BTC, the biggest.

Hit rate and repetition did the work. Not a fat reward-to-risk ratio.

What this record does not support: the idea that a good trader must cut losses fast and hold winners for a multiple of risk. This one did close to the reverse and still compounded.

What it does support: being right more often than not, on a market you trade thousands of times, with size you can survive being wrong on.

Method note: these are exchange-recorded realised-PnL events, not reconstructed round trips, and their sizes differ. Period: March 2018 to December 2021. Computed by us from his published files. Not investment advice.

Source (his own release post): https://gall.dcinside.com/mgallery/board/view/?id=chartanalysis&no=5051684

#Bitcoin #TradingData #RiskManagement #BitMEX